wasm scan demo: fetch layout into ./models/ via download_dependencies.sh in www/

Point the same-origin layout paths at ./models/ and document running the
installer from crates/docling-wasm/www/ — it writes ./models/ relative to
the current directory, so the page reads ./models/layout_heron_int8.onnx
(fallback ./models/layout_heron.onnx) with no copy step. gitignore the
www/models/ tree the installer populates.

Refs #157

Signed-off-by: artiz <artem.kustikov@gmail.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EY5KAiquN4YpVf2PXEQkVT

Author: artiz

.gitignore

@@ -8,6 +8,8 @@ onnxruntime_profile_*.json
 
 # wasm-bindgen output for the docling-wasm demo (built from source, see its README)
 crates/docling-wasm/www/pkg/
+# ML models the scan demo fetches same-origin (download_dependencies.sh, run in www/)
+crates/docling-wasm/www/models/
 
 # Secrets — never commit (holds CARGO_REGISTRY_TOKEN)
 .env

crates/docling-wasm/www/scan.html

@@ -48,22 +48,27 @@ <h1>docling.rs → wasm: scanned PDF / image, full lite profile</h1>
     reconstruction (TableFormer is stage 3). Nothing leaves this page.
   </p>
   <details id="setup">
-    <summary>The layout model is served from this page's own origin — fetch it once (~68&nbsp;MB int8)</summary>
+    <summary>The layout model is served from this page's own origin — fetch it once into <code>./models/</code></summary>
     <p>
       Unlike the recognition model (fetched from Hugging Face, which sends CORS
       headers), the RT-DETR layout export lives on this repo's GitHub Release,
       and <strong>GitHub Release assets can't be fetched cross-origin from a
       browser</strong> (no <code>Access-Control-Allow-Origin</code>). So this
-      page loads the layout model <em>same-origin</em> — put it next to this
-      file and serve the directory:
+      page loads the layout model <em>same-origin</em>, from a
+      <code>models/</code> directory next to it. Run the installer <strong>from
+      this <code>www/</code> directory</strong> — it writes <code>./models/</code>
+      relative to the current directory:
     </p>
-    <pre>curl -fsSL https://raw.githubusercontent.com/docling-project/docling.rs/master/scripts/install/download_dependencies.sh | sh
-cp models/layout_heron_int8.onnx crates/docling-wasm/www/   # ~68 MB int8 (or layout_heron.onnx, ~172 MB fp32)</pre>
+    <pre>cd crates/docling-wasm/www/
+curl -fsSL https://raw.githubusercontent.com/docling-project/docling.rs/master/scripts/install/download_dependencies.sh | sh</pre>
     <p>
-      These are the exact models the native pipeline uses. The page prefers the
-      int8 export and falls back to fp32; both decode identically here (same
-      <code>pixel_values</code>&nbsp;→&nbsp;<code>logits</code>,&nbsp;<code>pred_boxes</code>
-      contract).
+      That populates <code>./models/</code> (this page reads
+      <code>./models/layout_heron_int8.onnx</code>, ~68&nbsp;MB, falling back to
+      <code>./models/layout_heron.onnx</code>, ~172&nbsp;MB — int8 needs a
+      release that hosts it, otherwise fp32 is fetched). These are the exact
+      conformance-validated models the native pipeline uses; both decode
+      identically here (same <code>pixel_values</code>&nbsp;→&nbsp;<code>logits</code>,&nbsp;<code>pred_boxes</code>
+      contract). The recognition model still streams from Hugging Face on demand.
     </p>
   </details>
   <div id="drop">drop a scanned PDF or image here, or click to pick</div>
@@ -88,13 +93,14 @@ <h1>docling.rs → wasm: scanned PDF / image, full lite profile</h1>
     pdfjs.GlobalWorkerOptions.workerSrc =
       "https://cdn.jsdelivr.net/npm/pdfjs-dist/build/pdf.worker.min.mjs";
 
-    // The layout model is loaded SAME-ORIGIN (see the setup note above): GitHub
-    // Release assets carry no CORS header, so a browser fetch of the release URL
-    // fails — the model has to sit next to this page instead. Prefer the
+    // The layout model is loaded SAME-ORIGIN from ./models/ (see the setup note
+    // above): GitHub Release assets carry no CORS header, so a browser fetch of
+    // the release URL fails — download_dependencies.sh, run from this www/
+    // directory, drops the models next to this page instead. Prefer the
     // conv-only static-INT8 export (~68 MB, a big first-load win over the
     // ~172 MB fp32) and fall back to fp32; both share the same interop
     // contract, so the wrapper below is unchanged.
-    const LAYOUT_PATHS = ["./layout_heron_int8.onnx", "./layout_heron.onnx"];
+    const LAYOUT_PATHS = ["./models/layout_heron_int8.onnx", "./models/layout_heron.onnx"];
     const SCALE = 2.0; // px per PDF point — the native pipeline's RENDER_SCALE
 
     const meta = document.getElementById("meta");
@@ -214,7 +220,7 @@ <h1>docling.rs → wasm: scanned PDF / image, full lite profile</h1>
     const layoutSession = await loadLayout();
     if (!layoutSession) {
       document.getElementById("setup").open = true;
-      ready("layout model not found next to this page — see the note above");
+      ready("layout model not found under ./models/ — see the note above");
       throw new Error("no layout model available");
     }
     await recFor(lang.value);